Doina Precup is a Romanian-Canadian computer scientist and professor. She is a professor at McGill University, a core member of Mila (Montreal Institute for Learning Algorithms), and the head of the Montreal research team at Google DeepMind.
Temporal Abstraction and the Options Framework
Precup earned her Ph.D. from the University of Massachusetts Amherst in 2000, under the supervision of Richard Sutton. Her research pioneered the **options framework** in reinforcement learning. The options framework mathematically models "temporal abstraction"—allowing an AI agent to plan and make decisions over multiple steps and timescales, which is crucial for solving complex, long-term tasks.
Mila and McGill Leadership
Precup is a leading voice in Montreal's thriving AI hub, researching AI ethics, reinforcement learning safety, and medical applications of AI, particularly in monitoring patient vitals and neurodegenerative diseases.
